Multi-objective optimization of multiple impinging jet system through genetic algorithm

Abstract

In this study, a pneumatic contactless levitation system utilizing impinging jet arrays is investigated numerically and the optimum design set is proposed using a multi-objective genetic algorithm known as elitist Non-Dominated Sorting Genetic Algorithm (NSGA-II). Steady and turbulent impinging jet flow is characterized by jet Reynolds number in the range of 3000-15000. Preliminary validation tests have been carried out for multiple impinging jet flow to establish an accurate numerical model. In order to assess enhancements in the contactless levitation system, levitation capacity and energy consumption are selected as objective functions depending on the optimization parameters. Design parameters are nondimensionalized distance between sequential jet nozzles (2 < s/Djet < 6), the aspect ratio of the cylindrical pillars (1 < ARP < 5) and the jet Reynolds number. Cylindrical pillars are implemented to the original design to prevent jet-to-jet interaction and to maintain a passive flow control. The optimization algorithm is performed through 20 generations with a population number of 50. Optimum solution set represented by Pareto Front suggests trade-off designs between the objective functions which provide useful results for the design of the contactless conveying system. The optimum design points are observed at uniformly distributed jet Reynolds numbers, a singular optimum solution for s/Djet=2 and two unique solutions for ARP as 1.6 and 4.2.